Advice to decision maker on coal mining project

IESC 2014/041:Teresa Coal Mine(EPBC 2011/6094) –New Development

Requesting agency / The Australian Government Department of the Environment
The Queensland Department of Environment and Heritage Protection
Date of request / 05 February 2014
Date request accepted / 05 February 2014
Advice stage / Assessment

Advice

The Independent Expert Scientific Committee on Coal Seam Gas and Large Coal Mining Development (IESC) was requested by the Australian Government Department of the Environment andthe Queensland Department of Environment and Heritage Protectionto provide advice on the Teresa Coal Project in Queensland.

This advice draws upon aspects of information in the Draft Environmental Impact Statement (EIS), together with the expert deliberations of the IESC. The project documentation and information accessed by the IESC are listed in the source documentation at the end of this advice.

The proposed project is located approximately 17km north of Emerald and 70km south of Clermont in the Bowen Basin, Central Queensland. The project area is defined by two Mining Lease Application areas (MLA). MLA 70405 is approximately 7,000 hectares and covers the southern part of the project area while MLA 70442 is approximately 3,000 hectares and covers the northern part of the project area.

The proposed project is targeting the Corvus II coal seam within the German Creek Coal Formation of the Blackwater Group in the Bowen Basin. The proposed new underground coal mine and associated above ground infrastructure would produce up to eight million tonnesof run-of-mine (ROM) coal per year. The principal method of mining would be retreat longwall extraction, with a small section by partial block extraction. The life of the mine is anticipated to be up to 30 years. The current coal resource within the Corvus II coal seam is estimated at 310 million tonnes. The MLA is bound to the south and west by Theresa Creek and Gordonstone Creek to the northeast; tributaries of the Nogoa River, located to the south. Construction of the proposed project is anticipated to commence in late 2014 subject to mining lease and environmental approvals. The majority of construction works are anticipated to be completed by 2017, with full operation to commence in 2019.

The IESC would like to highlight thatthere are factual inconsistencies and key findings within chapters of this EIS that are not consolidated or considered within other chapters. For example conclusions and scenarios from one report (e.g. geotechnical findings of subsidence fracturing into the Tertiary formations) have not been considered in otherreports (e.g. the hydrogeological model). This leads to uncertainty regarding the impacts as predicted by the proponent. Notably the IESC considers the following to be key impacts of concern for the proposal:

  • Exclusion of subsidence induced fracturing into the Tertiary sands in the numerical groundwater model may under-predict mine water inflows and groundwater drawdown in this important regional aquifer.
  • Given there may be an underestimate of mine water inflows, discharge volumes would also be underestimated. A potential increase in this volume of water has not been assessed for impacts to flow regimes, geomorphology, water quality and consequential impacts on water related assets.
  • The impact of fracturing on water quality from the mixing of surface water and water drained from overlying aquifers has not been assessed nor has potential generation of acidic water from water percolation through potentially acid forming material above the goaf. This water will need to be removed from the mine and discharged to surface watercourses.

The IESC, in line with its Information Guidelines1, has considered whether the proposed project assessment has used the following:

Relevant data and information:The following pieces of relevant data and information are needed for potential impacts arising from proposal to be assessed:

  • Expanded spatial and temporal groundwater monitoring data to determine the behaviour of groundwater levels;
  • Analysis of subsidence fracturing impacts to surface watercourses, groundwater drawdown and mine inflows;
  • Data characterising the geomorphology of surface watercourses and surface water-groundwater interaction;
  • Data on the timing, location and water quality of surface water discharges;
  • Information related to water-related ecological values; and
  • A quantitative assessment of the proposal's contribution to cumulative impacts within the region.

Appropriate methodologies which have been applied correctly: Clarification and justification of the methods and approaches used to predict potential impacts are needed, particularly in relation to:

  • The conceptual and numerical groundwater models not sufficiently identifying impacts on water resources or other matters of national environmental significance. Notably this includes the exclusion of a model scenario where subsidence fracturing is present in the Tertiary units;
  • Incorporation of Tertiary sands into the subsidence modelling;
  • While flood modelling techniques are sound, accuracy and confidence in the model would be improved by adopting more recent analytical methods and providing further information to clarify and justify model inputs and results. The flood model, as compiled, is not suitable for assessing water quality impacts;
  • Inclusion of changes to floodplain flow distributions and velocities, and consequential impacts, on the rising and falling limbs of flood events; and
  • The use of mean monthly discharges to assess allowable water discharges implies that Theresa Creek is a perennial rather than ephemeral watercourse.

Reasonable values and parameters in calculations: Justification and/or further information is needed to support the proponent’s approach or conclusions in relation to:

  • The extent of subsidence fracture height and enhanced permeability causing an underestimation of ground and surface water impacts;
  • The selection of watercourse model boundaries in the groundwater model used to represent surface water-groundwater connectivity in the proposal;
  • The absence of formations in the subsidence model; and
  • Exclusion of faulting from the groundwater model.

TheIESC's advice, in response to the requesting agencies' specific questions is provided below.

Question 1: The proponent states in Volume 1; Chapter 13 of the EIS (Attachment A), that given the distance from mines in the region and relative geological isolation from these mines (except via the basalt strata), significant cumulative impacts are unlikely. Taking into account the information provided in the EIS regarding other groundwater users in the area, does the Committee agree that significant cumulative impacts are unlikely? If not, what information is needed from the proponent to support their claim?

  1. The IESC notes the absence of data presented in the EIS to enable anassessment of the likelihood of significant cumulative impacts to ground and surface water resources including the following:
  2. There are three mines located 30-50km east of the proposed project. A qualitative approach was adopted to assess cumulative groundwater impacts in the region. The IESC considers that this qualitative approach contains assumptions, regarding the size of the projects, drawdown impacts, timing of mining and drawdown,that may result in an underestimation of the impacts. The IESC recommends that the numerical groundwater model be updated to address the matters discussed in question 2 and that the results beused in the cumulative impact assessment.
  3. For groundwater users, the numerical model estimates extraction for irrigation as 2 ML/year for both stock and domestic bores and apportions all licensed extraction volumes equally across all bores associated with any given licence. If the actual extraction volumes vary across the model areas,then this assumptionmay underestimate drawdown associated with groundwater user bores and Groundwater Dependent Ecosystems (GDEs) mapped in those specific areas.
  4. Prior to construction, propertieswith potential drawdown of greater than 1m will have a baseline assessment undertaken at each of the active bores to establish pre-operational condition. As the current model may beunderestimating groundwater drawdown and groundwater impacts may be greater than those currently predicted, baseline monitoring should be extended to include those groundwater bores affected.
  5. The following features would enable a comprehensive assessment of the cumulative impacts on surface water resources in the Nogoa River catchment:
  1. Identification and sensitivity of potentially impacted ecological and human assets;
  2. The quality, timing and volume of water discharges for the proposal, and other mines operating in the Nogoa River catchment;
  3. Effect of cumulative water discharges on watercourse flow regimes and volumetric discharges, water quality, geormorphology, aquatic biota and ecosystems;
  4. Effects of cumulative groundwater drawdown on surface water hydrology and water quality;
  5. Information on increased surface water-groundwater connectivity as a result of subsidence-induced fracturing and cracking, and the consequential effect on regional surface water hydrology;
  6. Regional changes to floodplain storage and/or flood behaviour, and;
  7. Impacts of the proposed haul roads on overland flow paths, flood behaviour and water quality.

Question 2: The EIS (Attachment A), Volume 1; Chapter 13 and Appendix I presents information regarding groundwater impacts, studies and mitigations. Please advise if:a. The numerical and conceptual groundwater models presented are adequate to predict the impacts on groundwater. If not discuss what information would address this. b. The numerical and conceptual groundwater models are adequate or do they present any concerns that may impact upon water resources or any other matters of national environmental significance?c.The Committee agrees with the groundwater level recovery time predicted by the proponent?

  1. The conceptual and numerical groundwater models do not sufficiently identify impacts on water resources or other matters of national environmental significance. There are a number of inconsistencies noted in the numerical and conceptual groundwater model which reduce confidence in the predicted impacts on groundwater. These include:
  1. Modelled scenarios and rationale: The groundwater model does not incorporate zones of enhanced permeability in the Tertiary sediments which would be consistent with the subsidence predictions presented in the subsidence report. Groundwater modelling results are only presented for fracturing in the overburden material. Constraining enhanced permeability mayfurther underestimate mine water inflows and subsequent drawdown of groundwater. This couldresult in loss of surface water to groundwater and volumes of water for surface water discharge. Results of a modelled scenario consistent with the subsidence modelling should be presented to allow for a better estimate of the potential groundwater drawdown and discharge volumes.
  2. The Emerald Formation is noted to comprise undifferentiated deposits of several sediments (gravels, sand, sandstone etc),thoughthe proponent states that the plasticity of the Emerald Formation is likely to provide mitigation against increased permeability. However, if subsidence impacts propagate through the overburden, groundwater impacts will be larger than currently predicted which would need to be considered given significance of the alluvium, Tertiary Basalt and Tertiary sands as a water resource in the region.
  3. Boundaries:The river bed has been set equal to the river stage, meaning that the river boundaries allow baseflow out of the aquifer, but do not allow leakage from the watercourse to the aquifer. Thisconstraint does not provide fortimes between seasonal flows in the watercourses where groundwater may be recharged. Further, the numerical model sets alluvium discharges predominately into creeks. Given the project area’s climate, the ephemeral nature of watercourses and groundwater monitoring bore levels in the alluvium adjacent to the creeks, these constraints result in an inaccurate representation of natural ground and surface water interactions. Characterisation of seasonal trends in groundwater levels and interactions with surface water would provide a more adequate assessment.
  4. Recharge:Groundwater recharge is reported to be minimal, with the evidence forthis statement based on limited monitoring data collected over a period of below average rainfall. An estimate of groundwater recharge was undertaken using the water table fluctuation method. The model utilises parameters at the low end of recharge values.Given the relatively high hydraulic conductivity in the Tertiary sequences, recharge rates may be considerablyhigher, particularly within the alluvium and basalt.
  5. Conceptualisation:The region’s geological and hydrogeological characteristics have been represented as seven layers in the proponent’s conceptual groundwater model. While the main water resource units (alluvium, Tertiary basalt and Tertiary sands) have been presented as distinct hydrostratigraphic units, it was noted that several formations were combined, in particular the Emerald Formation and Permo-Triassic overburden. This may result in an over simplification of the hydrogeological regime, in particular the characterisation of the Permo-Triassic overburden and its role in mitigating potential subsidence impacts to shallower Tertiary aquifers.
  6. The proponent notes the existence of a fault at the western edge of the longwall panelsin the Permian strata, but this fault is not included in the model.Given the fault's proximity to the subsided area and Theresa Creek, the fault should be characterised so as to minimise uncertainty of its role in groundwater connectivity.
  7. The proponent notes the likely variability in extent of the Tertiary sands. The distribution, extent andhydraulic connectivity with units above the Tertiary sands need to be characterised, as it has been identified as a significant water resource in the region. Subsidence fracturing propagated through the Permo-Triassic overburden is likely to significantly impact on the Tertiary sands. Should subsidence fracturing induce flow from the Tertiary Sands, mine inflow will increase and there may also be a reduction in groundwater levels. This will impact users of this resource, particularly when recharge is not sufficient to offset the effects of drawdown.
  8. The proponent's interpretation of groundwater flow in the Permian strata does not appear consistent with conceptualised geology in the area. Groundwater flow is depicted to the north in the Permian, which is noted to be opposite to the dip of the Corvus 2 seam. This findingshould be interpreted with caution due to vast spacing of the monitoring bores. Further monitoring bores could provide clarification and justification of these interpreted contours.
  9. Model prediction: There was insufficient data to undertake transient model calibration. As the model is only calibrated in steady state it is likely to produce transient predictions of low confidence. The predictive model includes stresses that are outside the range included in calibration, such as the fracturing and dewatering due to subsidence.As such,the reliability ofand confidence in the model's predictions is likely to be low.
  1. For these reasons, the IESC considers that there is uncertainty in the current model’sconceptualisation, numerical construction and predictions, which may significantly underestimate groundwater impacts and volumes of surface water discharge due to mine water inflows.
  2. The proponent states“a period of approximately 15 to 20 years after closure is required for water levels to return to within 1m of final static levels”. Given the concerns relating to the model, the IESC considers that there is low confidence in the predicted groundwater level recovery.

Question 3: Does the Committee agree with the findings in Volume 1, Chapter 6 and Chapter 9 and Volume 2, Appendix F1 to F5 of the EIS (Attachment A) that changes to topography and hydrology as a result of subsidence to the west of the Blair Athol railway line are not likely to result in significant impacts during flooding from Teresa Creek? Does the committee find the proposed subsidence mitigation and management measures suitable to protect downstream surface water quality and quantity?

  1. The proponent’s flood study refers to a ‘TC5’ subsidence scenario; however, this scenario is not discussed in the EIS or supporting technical subsidence report. The flood study may be based on an earlier, and possibly different, prediction of subsidence effects. Confidence would be improved by confirmation that the flood study is based on the most recent subsidence predictions.
  2. Changes in peak flood surface levels will be accompanied by a change in flow volume, timing,extent, and/or depthentering and leaving the floodplain. Such a change has not been quantified. Alterations to flood volume could impact environmental assets or other beneficial uses downstream of the proposed development. Presenting and comparing the pre- and post-development flow hydrograph would inform the assessment of potential impacts on sensitive assets.
  3. The modelling techniques used to assess flood impacts appear to be sound; however the accuracy of the model and design outcomes would be improved by using more recent flood frequency analysis techniques, such as those developed by Rahman et al (2012) and Kuczera & Frank (2006), to reduce uncertainty in design flood flow estimation.
  4. Confidence in the flood models and the conclusions of the flood study would be enhanced by:
  1. Providing a comparison of pre- and post-development flood hydrographs and longitudinal profiles of peak flood surface along the floodplain within the mining lease;
  2. Providing the rationale for the use of the CRC Forge method to estimate design rainfall for flood levels greater than the 100 year average reoccurrence interval (ARI), rather than the Bureau of Meteorology’s Intensity-Duration-Frequency data;
  3. Resolving the discrepancy between the different peak flow rates generated for the 100 year ARI event and describing the method used to derive the adopted peak flow rate of 8,521 m3/s;
  4. Plotting measured annual series stream gauge data against the adopted flood frequency distribution to demonstrate the ‘goodness of fit’ of the adopted probability distribution;
  5. Undertaking and documenting a quality assurance process for model inputs, particularly in relation to the light detection and ranging (LIDAR) dataset;
  6. Quantifying the overall uncertainty in the model results. The models in their current form cannot be considered as calibrated to historical data. The sensitivity analysis should be expanded to include hydrological (initial loss, continuing loss, routing parameters, rainfall volume) and hydraulic (roughness and head loss coefficients) parameters that elucidatethe uncertainty range in the flood behaviour.
  1. The potential for increased surface water-groundwater connectivity in the Theresa Creek floodplain overlying the underground mine has not been considered. While subsidence-induced fracturing in the Theresa Creek floodplain is not expected to significantly impact peak flood conditions for 50 year ARI, or larger, flood events, the influence of fracturing on smaller floods may be more significant. Expansion of the flood study to incorporate groundwater flux and a broader range of flood events would enable the impact of groundwater drawdown on peak flood levels and flood volumes to be more comprehensively assessed and quantified.
  2. Flood flow velocity impacts have not been assessed by the proponent. Subsidence within the Theresa Creek floodplain is unlikely to significantly alter peak flood flow velocities during the 100 year ARI, or greater, flood event. Floodplain flow distributions and velocities are likely to be impacted on the rising and falling limbs of the flood, particularly as the flood wave tracks across the floodplain on the rising limb of the flood hydrograph. The potential impacts of changes to flood flow velocity are likely to be limited to changes in the mobilisation of sediments. Quantification of changes to flood flow velocities, incorporating rising and falling limbs of the flood event, as well as peak flows, would enable the full range of potential flood impacts to be identified and assessed.
  3. Due to the low gradient of the floodplain and relative depth of the subsided panels, the proponent has assessed that mitigation measures to prevent flooding of subsided parts of the floodplain, or drain ponded water from the floodplain, are not feasible. Although the construction of levees and diversion channels and drainage of ponded areas is possible, the impacts of implementing these mitigation measures may include changes to flood behaviour, geomorphology, and water quality, with consequential impacts on downstream water-related assets. An assessment of the benefits and potential impacts of flood mitigation measures is needed.
  4. The flood models, as compiled,are not suitable to directly assess water quality impacts. Water quality modelling would assess changes towater quality components. A finer scale grid resolution would be needed in the model to assess effects of micro scale surface erosion issues, such as subsidence-induced surface cracking or changes to velocity profiles in the subsided areas.

Question 4: Is flood modelling presented in the EIS (Attachment A), Volume 1, Chapter 9 and Volume 2, Appendices F1 and F3 sufficient to determine if flood mitigations and water management systems will be effective to prevent downstream impacts on water quality and quantity? If not can you recommend any improvements?